SourceJul 2, 20262026-07-024 minVerified
Rethinking Human Oversight for High-Risk Autonomous AI
A look at high-risk AI oversight through the humans-as-handlers approach, focusing on intervention, accountability, and trust.
CommunityJul 2, 20262026-07-024 minVerified
How Safety Holds Up Across Long AI Conversations
Examines whether AI safety remains consistent in long conversations and highlights gaps in session-level evaluation.
CommunityJul 1, 20262026-07-015 minVerified
AI Employment Narrative Shifts From Loss to Redesign
Examines whether AI eliminates jobs or redesigns tasks, and why this shift matters for hiring, reskilling, and productivity.
SourceJul 1, 20262026-07-015 minVerified
Design Axes for Agentic Orchestration in Enterprises
A practical guide to balancing agent autonomy, traceability, and control in enterprise orchestration design.
SourceJul 1, 20262026-07-014 minVerified
Emergent Misalignment Depends on More Than Training Data
EM may depend on optimizers and batch settings, making finetuning recipes part of safety evaluation, not just data.
SourceJul 1, 20262026-07-015 minVerified
How Agents Should Help Users Form Preferences
Why AI agents must move beyond preference elicitation to support preference formation, with evaluation and safety in view.
SourceJul 1, 20262026-07-014 minVerified
LLM Agents Split Research Ideas And Verification
How a dual-agent LLM pipeline separates proposing tighter relaxations from verification in automated research.
SourceJul 1, 20262026-07-014 minVerified
Rethinking AI Tutors Beyond Cloud Chatbots in Education
In education, AI design matters more than raw performance, with student privacy, data minimization, and teacher control at stake.
CommunityJul 1, 20262026-07-013 minVerified
What AI Pricing Hides About Safety Operations
Commercial APIs and open-weight models differ not just in performance, but in who runs blocking, logging, and policy enforcement.
SourceJun 30, 20262026-06-303 minVerified
AI Paper Review Between Assistance and Official Evaluation
Examines Google PAT's paper-checking results and limits, and where AI should fit in academic review workflows.
SourceJun 30, 20262026-06-304 minVerified
Choosing Conservative Equilibria in Offline Multi-Agent Reinforcement Learning
Why equilibrium selection, conservatism, and data coverage matter when solving offline multi-agent games from fixed logs.
CommunityJun 30, 20262026-06-304 minVerified
Deploying AI for Rhythm Games by Function
Rhythm game AI works best when API and local inference are split by function, balancing latency, limits, cost, and memory.
CommunityJun 30, 20262026-06-304 minVerified
How To Read AI Firms Calling For Regulation
How to assess whether AI firms' calls for regulation signal safety commitments, competitive strategy, or both.
SourceJun 30, 20262026-06-304 minVerified
Smaller Fast Weights Beat Bigger LSTMs in Traffic Forecasting
A compact fast-weight recurrent model reported lower pooled RMSE than a larger LSTM using only 22.4% of the parameters.
NewsJun 30, 20262026-06-304 minVerified
Why Humanoid Office Robots Need Real Validation
Office humanoid robots should be judged by learning pipelines, generalization, and public validation, not demos alone.
SourceJun 29, 20262026-06-294 minVerified
Beyond PR Passes: Governing Repositories for Coding Agents
Autonomous coding agents should be evaluated beyond PR pass rates, with repository-level risk and structural health in view.
SourceJun 29, 20262026-06-294 minVerified
Class Frequency Guided Noise Schedules in Diffusion Models
Examines how class imbalance affects score learning in diffusion models and why frequency-guided noise schedules matter.
CommunityJun 29, 20262026-06-293 minVerified
Cloud LLM Costs Versus Local Deployment Decisions
Compare cloud token-based LLM pricing with local deployment to assess cost, control, latency, and break-even conditions.
SourceJun 29, 20262026-06-294 minVerified
CoIn Rethinks 3D Scene Editing Without Precise Masks
CoIn links 2D inpainting and 3DGS to reduce reliance on precise multiview masks in 3D scene editing workflows.
CommunityJun 29, 20262026-06-294 minVerified
Do Language Models Really Build Stable World Models
Strong language performance may not imply a stable world model. Reassessing LLMs through failures in time, space, and physics.
SourceJun 29, 20262026-06-293 minVerified
GRACE Rethinks VLM Quantization With QAT and Distillation
How GRACE combines QAT and distillation to balance accuracy and deployment cost in vision-language models.
SourceJun 29, 20262026-06-293 minVerified
Measuring Domain Gaps in Cross-Sensor Diffusion Super-Resolution
Why top satellite SR models on synthetic data may not lead on real cross-sensor imagery, and how to evaluate the gap.
SourceJun 29, 20262026-06-293 minVerified
MMG-Pop Rethinks Social Popularity Prediction Across Platforms
MMG-Pop uses multimodal and temporal graph signals from Bluesky and Reddit to reassess social popularity prediction.
SourceJun 29, 20262026-06-294 minVerified
Ontology-Guided KGQA Cuts Noisy Multi-Hop Reasoning Paths
How ontology constraints reduce noisy paths in multi-hop KGQA and improve reasoning for complex queries.